Hello,
Exchange 2010
4 mailbox servers in a DAG, two on each site.
2 CAS Arrays consisting of 4 CAS servers, two on each site. We’re using an HLB on each site for the CAS Arrays.
4 Windows 2008 domain controllers - two on each site
The problem started when we powered down half our exchange environment and two of our domain controllers so that they could be moved to our new off-site location.
We did this in a controlled manner and ran all necessary maintenance scripts on exchange, obviously making sure all the databases were active on the remaining mailbox nodes.
We edited the DNS entry for the SITE 2 CAS array to point to the SITE1 CAS array IP address.
This ensured anyone that had their mailbox on a database that had SITE2 CAS array set as the endpoint would still work. It was just as easy to do it this way instead of changing the mailbox database endpoints.
Everything went smoothly until we started to receive a number of calls from users reporting that there was a delay on e-mails being delivered (from e-mails being sent internally and externally). Some users would receive delayed e-mails in batches. In some cases the delay was up to an hour, but in most it was around 20-30 mins. Clicking update folder in outlook every so often is one way to get around the issue.
All our users are using outlook 2010 in cached exchange mode.
When we turned off cached exchange mode there was no delay at all, so outlook had no problem reading the e-mails directly from the mailbox on the exchange server. Also, all blackberry devices received e-mails pretty much instantly.
With a number of users, we tried recreating the outlook profile to generate a new OST file. This did the trick in some cases but the problem still continued with most.
Next week we will be powering back on our other site.
I’d like to know what’s causing this delay issue. Can anyone shed some light?